breakdown of the substrate to small dextrins observed on one extreme with taka amylase and the relatively slow disappearance of long chain saccharides observed with swine pancreatic amylase on the other extreme appear to be definite properties of these amylases. It is challenging to speculate about and to seek the cause for these differences in the action of two amylases, both of which appear to be so-called simple p r ~ t e i n s . ~ >38l ~ - ~ * NOTEADDED OCTOBER20, 1953-Since this paper was submitted for publication, two additional series of experiments have been completed with results that confirm and strengthen the conclusions. When all of the amylases in the comparison reacted with the substrate a t the same hydrogen ion activity, a t PH 7.2, the results were identical with those reported in Fig. 3 and in Table I for comparisons made when each amylase reacted with the substrate a t the p H most favorable to its action. Thew results show that the differences ohserved in the ac-
tion of the amylases are true differences and are not due to the influence of differences in the hydrogen ion activities of the substrate. An unfavorable hydrogen ion activity may decrease materially the concentration of active amylase but does not influence its unique action. It must be remembered that the comparisons given here for amylase action are independent of amylase concentration, Fig. 1. Similarly, results identical with those reported in Fig. 3, curve 1, were obtained when suitable volumes of swine saliva that had been held a t 100' for 5 minutes to inactivate the amylase were added to recrystallized swine pancreatic amylase. Thus, the substances that accompanied the unpurified swine salivary amylase had no influence on the action of recrystallized swine pancreatic amylase and presumably had no influence on the action of the swine salivary amylase they accompanied. These results give additional evidence that the differences observed above in the action of swine pancreatic amylase and of swine salivary amylase represent true differences in the specific mode of action of these two amylases. NEWYORK, N. Y.
[CONTRIBUrION FROM THE DEP.4RTMENT OF CHEMISTRY AND THE SCIEYCE RESEARCH INSTITUTE, OREGON STATE COI.T,EGE]
Conversion of Acetate and Pyruvate to Tyrosine in Yeast' BY RICHARD C. THO MAS,^ VERNONH. CHELDELIN, BERTE. CHRISTENSEN AND CHIHH. WANG RECEIVED NOVEMBER 22, 1952 CH&140COOH and CHaC1400Hhave been compared as carbon sources for the formation of tyrosine in bakers' yeast. The intramolecular distribution of radioactivity in the isolated tyrosine indicated that the aromatic ring was formed from pyruvate via oxalacetate or a similar unsymmetrical Ca-acid as intermediate. The side chain appeared to arise from pyruvate as an intact G u n i t .
Considerable interest has been focused recently upon the biosynthesis of tyrosine in microorganisms. Davis's studies3 with mutant strains of Escherichia coli point to shikimic acid as a normal precursor of tyrosine and phenylalanine, as well as other aromatic compounds in that organism. Baddiley, et ~ l . have , ~ suggested that when acetate is used as the sole carbon source for adapted Torula yeast, the carboxyl group gives rise to the carboxyl of tyrosine, as well as to carbon 4 of the ring; the other seven carbon atoms were considered derived chiefly from the methyl group of acetate. Finally, the side chain of tyrosine may arise from pyruvate as an intact unit in E. coli (Cutinelli, et aL6). In this Laboratory, CL4-labeledpyruvate and acetate have been compared as carbon sources for yeast growth. The radioactivity distribution patterns in the isolated tyrosine suggest that oxalacetate or other unsymmetrical C4acid may function as an intermediate in the conversion of pyruvate to this amino acid. Experimental Use was made of the yeast samples obtained previously,6
iii which 20 mmoles each of CH1C140COOHor CH3CI400H ( 1 ) This research was supported by contract No. AT(45-1)-301 from the Atomic Energy Commission. Published with the approval of the Monographs Publications Committee, Research paper no. 230, School of Science, Department of Chemistry. Presented before the Northwest Regional Meeting of the American Chemical Society, Corvallis, June, 1952. (2) National Science Foundation Predoctoral Fellow. (3) B. D . Davis, J . Biol. Chem., 191, 315 (1951). (4) J. Baddiley, G . Ehrensvard, E. Klein, L. Reio and E . Saluste, i b i d . , 183, 777 (1950). (5) C.Cutinelli, G.Ehreasvard, L. Reio, E. Saluste and R. Stjernholm, Acta Chem. Scand., 5, 353 (1951). (0) C . H. Wang. R F. T,ahhe, € 3 E Chri.;t?iiseii an4 V TI CIipIdcIiri, .I R i d Chrm , 197, R t R ;IqY,l.)
with a specific activity of 1.85 X lo8 c.p.m. per mmole, were administered as the sole carbon source to Fleischmann's bakers' yeast that had been previously grown on glucose. In the pyruvate experiments, all of the labeled substrate was utilized in 4 hours aerobically, and 5 hours anaerobically. Acetate was employed only under aerobic conditions, 39% of the labeled substrate being used in 4 hours. Details of these fermentations have been presented elsewhere.6 Tyrosine was isolated from the yeast hydrolysate7 by concentration and crystallization at the isolectric point, following a fivefold dilution with non-isotopic L-tyrosine. The specific activities were thus one-iifth of the values given in Table I. Yields of the (diluted) tyrosine, obtained from three grams of dry yeast in each sample, were: from acetate, 81.0 mg.; from pyruvate (aerobic), 91.2 mg.; and from pyruvate (anaerobic), 66.1 mg. Purity of the isolated samples was established by paper chromatography. The tyrosine obtained was degraded according to the method of Baddiley, et d.,* on the same scale in the following manner: (1)combustion t o COI for the specific activity of the whole molecule; (2) decarboxylation with ninhydrin for the specific activity of the carboxyl carbon; (3) fusion with KOH and NaOH t o give p-hydroxybenzoic acid. Combustion and radioactivity assay gave the specific activity of this compound directly and the specific activity of the amino carbon of tyrosine by difference; (4) nitration of phydroxybenzoic acid to 3,5-dinitro-4-hydroxybenzoicacid. The latter was oxidized with Ca(0Br)z t o bromopicrin which was in turn converted t o Con. This represented carbon atoms 3 and 5 of the tyrosine ring; (5) nitration of p-hydroxybenzoic acid t o picric acid. A small amount of the picric acid was burned to Cot t o obtain the specific aotivity of the benzene ring as well as the specific activity of the methylene carbon atom of the side chain by difference. The remainder of the picric acid was oxidized with Ca(0Br)z t o bromopicrin, which was oxidized t o Cot. This represented carbon atoms 1, 3 and 5 in the tyrosine ring. The average specific activity of carbon atoms 2 , 4 and 6 could be thus obtained by difference. Carbon atom 4 of the ring was not differentiated from (7) R . F. Labbe, R. C. Thomas, V.H. Cheldelin, R . R. Christenscri ;Ind C. H. Wang, J. B i d . Chem., 197, 655 (1952).
Nov. 20, 1958
CONVERSION OF ACETATEAND PYRUVATE TO TYROSINE IN YEAST
5,555
carbon atoms 2 and 6 since only a very small amount of tyrosine was available. The determination of the activity of the @carbonin the side chain by difference, instead of by direct decarboxylation of p-hydroxybenzoic acid,4 *was also due to the limited amount of sample available in the present experiments. All samples were counted as BaCOB with correction for background and self-absorptionin the conventional manner.
of pyruvate to glucose and cyclization of the latter, since this would produce only two active centers in the ring, a t para positions. A study of other possible metabolic pathways to tyrosine via shikimic acid revealed a good reconciliation of the observed radioactivities, through the use of two molecules of oxalacetate or other Cd-acids with similar isotopic patterns (formed from pyruResults and Discussion vate via C& condensationa) as the starting maThe radioactivity distribution in tyrosine from terials. Thus, in this speculative sequence (Scheme each of the three yeast samples is given in Table I. I), the hypothetical condensation product is cyAs observed previously for aspartic acid8 and glu- clized, aromatized, condensed with pyruvate (or its tamic acidIg the distribution patterns differed equivalent) and transaminated. When oxalacetate widely in the acetate and aerobic pyruvate sam- was assigned the radioactivity distributions obples. Also, as experienced before, the anaerobic served for aspartic acid in this yeastla the calcupyruvate sample exhibited an intermediate pattern lated activities within the ring agreed closely with of activity. those observed, as follows: with C1 of the ring TABLEI considered equivalent to CS of oxalacetate, the DISTRIBUTION OF C14 IN TYROSINE FROM YEASTUTILIZINGvalue of 8.33 X lo4 c.p.m./mmole carbon should be accompanied by counts of 11.8 X lo4 in C3 + 6, CH&YOOH OR CH3C140COOH Aerobic Anaerobic Aerobic and 4.5 X lo4 in CZ + 4 + 0. (Found: 11.8 and Conditions pyruvate pyruvate acetate 5.7 X lo4 in these positions, respectively). The C.p.m. Per C.p.m. Per C.p.m. Per cent. per cent. cent. per Pa' activity of CIof the ring could not, of course, be mmole of mmole of mmole of expected to equal that of the amino carbon, since x 104 total x 104 total X lo4 total the relative dilution of labeled pyruvate and CqWhole molecule 49.1 100 19.0 100 4.83 100 acids with cellular constituents was unknown 15 2.43 50 COOH 4.3 9 2.8 The present proposal that tyrosine may arise CHN'Ha 18.0 37 4.9 0 26 0 from unsymmetrical C4-acids, differs from the CHa 0 0 0 0 0 0 earlier conclusion12that this amino acid is produced CsH4OH 25.9 53 10.5 55 2.59 53 from glucose viu cyclization. Later work with gluDistribution of C14 within the ring cose as substrate" has ruled out direct cyclization, C~HIOH 25.9 100 10.5 100 2.59 100 and the authors have pointed out the possibility of c 1 8.3 32 1.6 15 0 0 condensation of a C3 unit with some other product cs+s 11.8 46 3 . 8 36 0.74 29 of glucose metabolism. However, the isotopic CZc4+8 5.7 22 5.0 48 1.74 72 data obtained did not permit a clear interpretation the course of tyrosine synthesis. C7 sugars were With pyruvate as substrate (aerobic fermenta- of considered as intermediates, based on the work of tion), 46y0 of the total tyrosine activity was located Benson, et al.,13 and Horecker and Smyrnioti~.'~ in the side chain, predominantly in the amino carbon atom. This atom presumably corresponded SCHEME I to the carbonyl carbon of the administered pyru- POSTULATED CONVERSION OF AN UNSYMMETRICAL G-ACID vate. This observation, together with the comTO TYROSINE plete absence of radioactivity in the /3-carbon atom, Values given in parentheses are for oxalacetate, in c.p.m. strongly suggested that the side chain of tyrosine X 10' per mmole of carbon. was derived from pyruvate as an intact unit. The (1.19) (0.36) (2.80) (0.80) 4 8 2 1 small amount of activity in the carboxyl group can HOOC-C-CCOOH probably be accounted for by its partial equilibraII tion with metabolic C02.6~'~ Origin of the tyro0 sine side chain from an intact C3 unit is also sug"2 gested by Cutinelli, et U Z . , ~ as well as by the data of 1C c-LCOOH Gilvarg and Bloch." In the latter, using glucosel-CI4as the carbon source, the &carbon of the tyrosine side chain contained nearly half of the total acI I tivity in the molecule. This carbon atom appar4 c n,caI etc. 4c, ently corresponded to the labeled atom of glucose Li 'L. after glycolysis. 1' Over three-fourths of the remaining radioactivVery recently, Shigeura, et al.,15 in a study of the ity (Le., in the aromatic ring) was found in positions 1, 3 and 5; CI was somewhat more active biosynthesis of shikimic acid (hence presumably also than the mean of CS and Cg. This distribu- tyrosine) in E. cog, have suggested that glucose is tion of activity cannot be explained by conversion not utilized per se, but instead undergoes fragmentation to produce two or more unsymmetrical units (8) C. H.Wang, R. C. Thomas, V. H. Cheldelin and B. E. Christen,
n / c 2 1
I
sen, J . Bioi. Ckcm., 197,663 (1952). (9) C. H. Wang, B. E. Christenen and V. H. Cbeldelin, i b i d . , 101,
683 (1953).
(lo)
0. EhrensvBrd, L. Reio, E. Saluste and R. Stjernbolm, ibid.,
189,93 (1951). (11) C. Gilvarg and K. Bloch, ibid.. 199,689 (1958).
(12) C. Gilvarg and K . Bloch, Tam JOURNAL, 71, 5791 (1950). (13) A. A. Benson, J. A. Bassham and M. Calvin, ibid., 71, 2970
(1961). (14) B. L.Horecker and 2. Smymiotis, ibid., 74, 2123 (1952). (15) H.Shigeura, D.B. Sprinaon and B. D. Davis, Faderalion Proc., ia, 459 (1953).
5556
JAMES
R. VAUGHAN, JR., A N D JOYCE A. EICHLER
which recombine to form the appropriate cyclic compounds. Both their work and the present study, in which the observed labeling patterns are compatible, demonstrate the need for intermediates that do not equilibrate with symmetrical products of glycolysis or the Krebs cycle. From pyruvate, this is more easily envisioned, since randomization or interaction with glycolysis products does not occur extensively; from glucose, the possible pathway is less clear. C4-acid formation through C1 C1 condensation is favored by pyruvate (as distinguished from acetate,Q) and the failure of tyrosine to be formed from C14-bicarbonatein the absence of pyruvate or other substratel6 may reflect the previously observed importance of the CrCl condensation reaction in yeast and its dependence upon pyruvate.8 The conversion of acetate to tyrosine may also proceed through the Cd-acids, although quantitative agreement between observed and expected values is less satisfactory than in the pyruvate sample. The high activity of the carboxyl group, as well as the high ring activity at conh the intramolecular distribution reported by Baddiley, et u Z . , ~ as far as the present degradation stud-
+
(16) L. Levy and M. J. Coon, J . Bioi. Chem., 192, 807 (1951).
[CONTRIBUTION FROM
THE
Vol. 75
ies were carried out. It may thus be assumed that Cr of the ring contains most of the activity designated in Table I as G+wa. This would be expected if oxalacetate were the precursor of tyrosine, since aspartic acid derived from acetate in this yeast contained C14 only in the carboxyl groups.8 It would also account for the absence of isotope in C1 of tyrosine. The deviation between expected and observed values occurs at C3+6;the activity of these atoms should equal the activity of C,. An inspection of the C14 distribution obtained by Baddiley, et aL4 (using C13H&1400Has substrate), reveals a similar observed deficiency of radioactivity at CWS,although the C13 distribution agrees well with expectations based on the foregoing scheme. The present results, as well as 0thers,'~-~~~'7 suggest that the citric acid cycle does not participate directly in the formation of tyrosine from pyruvate, acetate or glucose, although the over-all effect of Krebs cycle intermediates is to increase tyrosine synthesis by yeast.'* Studies are in progress to relate further the behavior of glucose and pyruvate in these transformations. (17) C.Gilvarg and K. Bloch, ibid., 193, 339 (1051). (18) A. Kleinzellet and G. Kubie, Chcm. Listr, 46, 106 (1952).
CORVALLIS, OREGON
CHEMOTHERAPY DIVLSION, STAMFORD RESEACH LABORATORIES, AMERICANCYANAMID COMPANY]
The Preparation of Optically-active Peptides Using Mixed Carbonic-Carboxylic Acid Anhydrides BY JAMES R. VAUGHAN, JR.,
AND JOYCE
A. EICHLER
RECEIVED JUNE11, 1953 The study of racemization by this method of synthesis has been continued. Using N-carbobenzoxyamino acid or peptideisobutylcarbonic acid anhydrides and amino acid or peptide esters, L-lysyl-L-valpl-L-phenylalanylglycine was prepared by two routes and shown to have almost identical properties in each case. Therefore, very little, if any, racemization occurred. When these mixed anhydrides in anhydrous solution were caused t o react with the sodium salts of amino acids (or peptides) in aqueous solution no racemization was observed if the mixed anhydride was formed from an optically-active N-carbobmoxyamino acid. Partial racemization was observed, however, if the mixed anhydride was formed from an optically-active N-acylamino acid, ;.e., an optically-active N-carbobenzoxy dipeptide, etc.
In a preliminary report' it was demonstrated synthetic material prepared here with his natural that, in the reaction employing mixed carbonic-car- product in the form of their dinitrophenyl (DNP) boxylic acid anhydrides for the formation of carbo- derivatives, and found them to be essentially idenbenzoxy peptide esters, racemization can be avoided tical.3 No conclusions concerning the optical idenor greatly minimized by the proper choice of reac- tity of the individual amino acids could be drawn, tion solvents and conditions. This investigation however. For comparison of optical purity, the derivatized has now been extended to the preparation of the tetrapeptide L-lysyl-L-valyl-L-phenylalanylglycinetetrapeptide was prepared by two routes. In the and to a study of the reaction of mixed carbonic- first of these, dicaxbobenzoxy-L-lysyl-L-vaSinewas carboxylic acid anhydrides with salts of amino prepared by saponificaton of its ethyl ester and acids or peptides under aqueous conditions. In- coupled through its mixed isobutylcarbonate anhyterest in the above-mentioned tetrapeptide was ini- dride with ethyl L-phenylalanylglycinate to give tiated by the announcement2 that the sequence on the dicarbobenzoxy tetrapeptide ester in 54% yield D i 0.6" (c 2, glacial the amino end of the single polypeptide chain of ly- m.p. 199-200°, [ c Y ] ~ ~-21.4 sozyme is lysylvalylphenylalanylglycyl-,and it was acetic acid). By the second route carbobenzoxy-L-valine was hoped that a direct comparison of a synthetic material with the natural product might be made. first condensed with ethyl L-phenylalanylglycinate This bas now been accomplished through the co- to give the tripeptide derivative, which was then operation of Dr. Walter A. Schroeder of the Cali- hydrogenated to ethyl L-valyl-L-phenylalanylglyfornia Institute of Technology, who compared the cinate. This was coupled with dicarbobenzoxy-Llysine by the standard methodl to give ethyl dicar(1) J. R. Vaughan, Jr., TEISJOURNAL, 74,6137 (1952). (3) W.A. Schroeder. i b i d . , 74, 5118 (1952). (2) W.A. Schroeder, ibid., 74,281 (1952).
